Protocol for the development of a tool (INSPECT-IPD) to identify problematic randomised controlled trials when individual participant data are available

IntroductionRandomised controlled trials (RCTs) investigate the safety and efficacy of interventions. It has become clear however that some RCTs include fabricated data. The INSPECT-SR tool assesses the trustworthiness of RCTs in systematic reviews of healthcare-related interventions. However, where individual participant data (IPD) can be obtained, a more thorough assessment of trustworthiness is possible. Consequently, INSPECT-SR recommends obtaining IPD to resolve uncertainties, though there is no consensus on appropriate methods for forensic analysis of raw data. Our aim is to evaluate IPD checks to establish which are worthwhile, and how they can be implemented in a new tool, INSPECT-IPD (Investigating Problematic Clinical Trials with Individual Participant Data). Methods and analysisUsing international expert consensus and empirical evidence, the INSPECT-IPD tool will be developed using five stages: (1) compiling a list of IPD trustworthiness checks, (2) evaluating the usefulness and ease of interpretation of the checks when applying them to a collection of presumed authentic and fabricated IPD datasets, (3) a Delphi survey to determine which checks are supported by expert consensus, (4) a series of consensus meetings for selection of checks to be included in the draft tool and finally (5) prospective testing of the draft tool in: a) the production of systematic reviews, and b) the journal editorial process for RCT submissions, leading to refinement based on user feedback. Ethics and disseminationThe University of Manchester ethics decision tool determined that ethical approval was not required (18 June 2024). This project includes secondary research and surveys of healthcare researchers on topics relating to their work. All results will be published as preprints and open-access articles, and the final tool will be freely available. STRENGTHS AND LIMITATIONS OF THIS STUDYO_LIAn international consensus process and empirical evidence will be used to develop the tool. C_LIO_LIThe development and dissemination of the tool will involve key stakeholders. C_LIO_LIIn the absence of a gold-standard test for problematic data, this tool should not be interpreted as a diagnostic instrument for trustworthiness. Instead, it will assist researchers in assessing the trustworthiness of a study. C_LIO_LIThe tool will only be applicable when individual participant data (IPD) can be accessed. Where IPD can be accessed, the ability to assess trustworthiness will be improved. C_LI